An Improved Parameter less Data Clustering Technique based on Maximum Distance of Data and Lioyd k-means Algorithm
| Autor: | Tutut Herawan, Khandakar Rabbi, Wan Maseri Binti Wan Mohd, A. H. Beg |
|---|---|
| Rok vydání: | 2012 |
| Předmět: |
Computer science
business.industry Correlation clustering Pattern recognition K-Means Algorithm Clustering Determining the number of clusters in a data set Data stream clustering Ramer–Douglas–Peucker algorithm CURE data clustering algorithm Canopy clustering algorithm Data Mining General Earth and Planetary Sciences Artificial intelligence Partitioning Clustering Algorithm business Cluster analysis General Environmental Science FSA-Red Algorithm |
| Zdroj: | Procedia Technology. 1:367-371 |
| ISSN: | 2212-0173 |
| DOI: | 10.1016/j.protcy.2012.02.076 |
| Popis: | K-means algorithm is very well-known in large data sets of clustering. This algorithm is popular and more widely used for its easy implementation and fast working. However, it is well known that in the k-means algorithm, the user should specify the number of clusters in advance. In order to improve the performance of the K-means algorithm, various methods have been proposed. In this paper, has been presented an improved parameter less data clustering technique based on maximum distance of data and Lioyd k-means algorithm. The experimental results show that the use of new approach to defining the centroids, the number of iterations has been reduced where the improvement was 60%. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|